Random BSOD when playing games

[haze11111]

Short version: My PC is getting a BSOD at random times when full screen gaming. The BSOD screen says the video card times out.

Long Version:
I have a gaming PC that a friend generously gave me when he built a new one. It's worked great, but in November the Tagan ITZ 700 watt power supply died. He had planned to do SLI and never did, so during Neweggs holiday sale I picked up a Corsair Enthusiast Series TX650 thinking it should be sufficient.

Everything has run smoothly with the new PSU except for the fact that when playing games like Torchlight 2 and Starcraft 2, my machine will BSOD. I had random BSOD's with Starcraft 2 and then suddenly had 4 nights in a row with no problems, but now it keeps crashing at the same spot in the campaign. There is little consistency to the crashes other than that and under normal operations, like email, web browsing, etc, it has only BSOD'd once and that was during a youtube video.

My first thought is that the one thing that has changed recently is the PSU and prior to that I only recall a rare crash here and there. I thought maybe my new PSU wasn't good enough, but everything I've read online about the requirements for the GTX 460, including the manual, say 450 watt minimum and I exceed that.

Here is a list of things I have tried:

- Updated video card and mobo drivers, I have also uninstalled the nvidia drivers, run Driver Sweeper in safe mode, and reinstalled drivers in normal mode. I have also tried legacy drivers.
- Ran GPU-Z to look at temperatures and ran OCCT to stress test the GPU and GPU Ram
- Ran memtest86+ for 12 hours, 8 passes
- Tried running the video card in another PCI-E slot
- Tried running games at lowest graphical settings with no success
- I ran Driver Verifier for 24 hours, but never got a BSOD. Maybe run it longer?
- I also found a link talking about the TDR errors with Windows 7 and added 6 items to the registry that were supposed to help give the video card 60 seconds to catch up, but Starcraft 2 still froze up, so I have since removed the registry items I added.
- Run error-checking via Windows 7 on both HDD
- I have not done any overclocking. I think my video card comes with some overclocked settings but in the bios the clock speed is the lowest it can go.

That's all I can think of right now, hope I didn't forget anything. Any help is most appreciated. Here are my system specs:

Processor: Intel Core 2 Duo CPU E8400 @ 3.00GHz
Mobo: MS-7380 (MSI P7N SLI Platinum)
Video Card: Palit GTX 460 Sonic Platinum
PSU: Corsair TX650 Enthusiast Series
RAM: 4GB
HDD: 80 GB Intel SSD
2nd HDD: 149GB SATA
OS: Windows 7 Home Premium 64-Bit 6.1.7601 Service Pack 1 Build 7601

 

[Puppies04]

Just to put your mind at rest, unless that PSU is faulty it will run your system with tons of headroom left.

 

[Puppies04]

Have you tried running afterburner and down clocking your GPU/GPU memory by 10% to see if the problems persist? (alternative would be raising the voltage but I wouldn't suggest that unless you nail the problem down with the previous suggestion) If you have a factory OC'd card it may just be on the verge of being unstable which can show these sorts of intermittent faults.

Also you mention running GPU-Z to check your temps but what were they under stress?

 

[haze11111]

Under stress tests the GPU only got up to 75C. I do believe that the Palit GTX 460 is a factory OC'd card, but it ran fine for me for many months before I had to swap out the power supply. I don't know of a way to down clock it, I checked the bios and the setting for the video card clock speed is as low as it will go.

I will try afterburner, never used it before so I will see if it can slow down the card. If the card was the problem though, wouldn't the stress tests bring it down? I ran tests with OCCT, Furmark, and 3Dmark 11. None of the tests ever caused a BSOD.

 

[mlc]

some of those Palit cards are know to have issues with heat.. but I don't recall the thresholds..

GIven you max at around 75 during stress, it's probably ok.. but I would be inclined to rule it out by pulling the card, cleaning up the fan and heatsink and reapplying some thermal paste.. then go in and change the fan profile to be more aggressive.. to see if that makes a difference.

 

[xilience]

Your best bet is the GPU because that component is the most stressed during gaming.

Take a stopwatch and see how long gaming lasts. (with relatively same scenario) This is a quick way to find out if it's simply overheating. If this is the problem you can try to better cool your system or possibly even look into aftermarket coolers for your GPU.

If this isn't the issue, it may actually be a driver problem. Search for an updated, and even an OLDER driver, as changes can sometimes have a negative affect as well.

Good luck.

 

[haze11111]

Would it still be a heat issue for the GPU if I ran a game with low settings? Example, right now the one consistent crash I can make happen is to load up the Starcraft 2 mission "Supernova" and the game will always crash right after the adjutant tells me the fire wave is coming. This occurs when using Ultra settings, but also "low" settings when the card is not stressed at all.

GPU fan and the case itself are recently cleaned up. I just wish I had an easy way to know if I was having a hardware problem or software. I'm debating wiping C: this weekend and reinstalling just windows, drivers, and games to see what happens.

Edit: I have tried older nvidia drivers from early this year, but no matter which driver I've tried I still crash in SC2 on that same mission. My other gaming crashes, like in Torchlight 2, are nowhere near as frequent. In over 20 hours of Torchlight 2 play it only happened twice.

 

[Puppies04]

Try afterburner, it should let you downclock the card to find out if that is the problem. It is more than possible a GPU can run stable for ages then suddenly become unstable as it ages. If you enable voltage control you also could try raising the voltage slightly to see if that cures the problem as I said before.

Heat could also be a factor, the hotter GPUs/CPUs get the more voltage they require to run stable.

 

[mlc]

if you're hitting a certain temp (e.g. 75c)..then it doesn't really matter what settings are doing it. .it's something to address.. Is that temp really high enough to cause an issue? who knows, probably not, but its close enough to investigate given all the other things you've tried and ruled out.

given that you've already did the clean & paste piece, then i would pursue altering the fan profiles to see if pushing more air onto the card during gameplay gets the temp lower, and possibly addressing your problem...

the other alternative would be to underclock , as suggested by puppies04, and to monitor the temps using that program.. to see if that helps...

 

[haze11111]

Ok I tried Afterburner and took the GPU clock and GPU ram clocks down to the minimum (405 GPU and 900 for the ram) I was able to play a few more minutes into the mission in Starcraft that keeps crashing me, but I still hit the same BSOD as before. Tried again and it happened right away. GPU temp was only 35C.

I am open to using Afterburner to raise the voltage, but what do you suggest raising it to?

My next step (if the increase in voltage does not work) is to try my old Radeon X80XL. That card is probably at min specs for SC2, but I will try it just to see what happens and to try and rule out my current video card. If that old card has the same BSOD then I might try taking out 3 of my 4 DIMMS and trying the game with just 1 installed and rotate through all of them to see if there is a bad stick. They made it through memtest, but I figure I will try anything to try and narrow down the problem.

 

[haze11111]

Latest Update:

I replaced my video card with my old ATI Radeon X800XL and was able to play that SC2 mission that was crashing my system. I had to play at a low resolution and with low detail, but there was no BSOD.

So, do you think I am right to think that it is either the video card or a defective PSU?

I just find it odd that my GTX460 + PSU can handle running the stress tests, yet, crash so often in SC2, or Torchlight 2, etc. I'm on a tight budget and don't know what to aim to replace, just bough that PSU in November when Newegg had a sale. Any thoughts on how I can narrow down the problem further?

Also worth noting, the old video card is powered by the PCI-E slot and a molex connector for an aftermarket fan while the GTX460 use two 6 pin power connectors straight from the PSU. That is why I am still thinking it could be a PSU problem because that is the one piece of hardware that has changed and my problems started after that.

 

[Steltek]

Latest Update:

I replaced my video card with my old ATI Radeon X800XL and was able to play that SC2 mission that was crashing my system. I had to play at a low resolution and with low detail, but there was no BSOD.

So, do you think I am right to think that it is either the video card or a defective PSU?

I just find it odd that my GTX460 + PSU can handle running the stress tests, yet, crash so often in SC2, or Torchlight 2, etc. I'm on a tight budget and don't know what to aim to replace, just bough that PSU in November when Newegg had a sale. Any thoughts on how I can narrow down the problem further?

Also worth noting, the old video card is powered by the PCI-E slot and a molex connector for an aftermarket fan while the GTX460 use two 6 pin power connectors straight from the PSU. That is why I am still thinking it could be a PSU problem because that is the one piece of hardware that has changed and my problems started after that.

If you bought it back in November, the power supply should still be under warranty. You could always contact Corsair and talk them about doing an RMA swap under warranty.

I presume that the video card is long out of warranty?

 

[nature1ders]

I find it unlikely to be the PSU. I think it's some other issue, my first guess here, hard drive? ok bye

 

[haze11111]

If you bought it back in November, the power supply should still be under warranty. You could always contact Corsair and talk them about doing an RMA swap under warranty.

I presume that the video card is long out of warranty?

The video card is long out of warranty. I thought about doing an RMA on the PSU, but, I still don't even know for sure if that is the problem. If I leave the computer on all the time and just surf the web, check email, etc, it never does a BSOD. This is also why I doubt it is an HDD problem and error checking turned up no problems there.

Any ideas on how I might narrow down whether it is the PSU or vid card?

 

[denis280]

RAM: 4GB

what brand.are they compatible??

 

[Steltek]

Is the motherboard BIOS up to date? You might also record the BIOS settings, reset the CMOS memory, and re-enter and save the BIOS entries. At least this would let you know if it is a BIOS glitch.

 

[haze11111]

what brand.are they compatible??

I'd have to dig out the DIMMS to check but they haven't changed since the computer was built in 2010. Over the years there have been zero problems until the power supply died and was replaced in November. If this is a ram issue then it is because they are damaged and I believe they are fine after memtest+ was run overnight for 12 hours, 8 passes, zero errors.

 

[haze11111]

Is the motherboard BIOS up to date? You might also record the BIOS settings, reset the CMOS memory, and re-enter and save the BIOS entries. At least this would let you know if it is a BIOS glitch.

I'll give this a go. I guess something could have happened to the BIOS when the old power supply died? I'll check bios version and update to the latest.

 

[haze11111]

I successfully updated the bios and also used MSI's utility to scan and update the MB drivers. Loaded up Starcraft 2 and was able to play through the mission that kept crashing and the mission after that.

However, on the third mission I tried to play I paused the game to take a phone call and while I was on the phone I saw it get the same BSOD.

Stop 0x00000116
nvlddmkm.sys

The description at the top was same as always, video card timeout.

Really frustrated. Just don't know what to do.

 

[haze11111]

Sorry to dredge up an old topic, but I had shelved this problem for awhile before finally finding a solution and thought I would post it here on the off chance someone else runs into something like this.

I changed out the surge protector.

I had an extra one laying around and decided to just try it and I haven't had any BSOD's ever since. I'm able to play games again no problem.

My guess is that when the PSU started to draw more power for the video card during games the surge protector was not able to supply it without fluctuation, hence, the crashes.

 

[Ketchup]

I have heard of issues like this with battery backups, but not many with a surge protector. Glad you figured it out.